Intel Core 5 120UL vs Intel Core Ultra 5 338H Comparison

The Intel Core 5 120UL and Intel Core Ultra 5 338H represent two very different Intel design points: a 10-core Raptor Lake desktop chip and a 12-core Panther Lake mobile chip. The recorded data is decisive. Across 17 head-to-head benchmark comparisons, the Core Ultra 5 338H wins every one. Its average benchmark score of 33989 places it in the 84th percentile of all CPUs, while the Core 5 120UL averages 13594 and sits in the 68th percentile. The 338H's nearest-rival deltas are -0.3% against the Core Ultra 7 165H and 0.3% against the Core i7-12800HX, placing it in the same performance band as those processors. The 120UL's nearest-rival deltas are 0.7% against the Core i3-12100F and 0.8% against the Core 3 N355.

Where Each One Wins

The win split is 0 for the Core 5 120UL and 17 for the Core Ultra 5 338H. No recorded benchmark favors the 120UL. The closest contest is Cinebench R23 single-core, where the 120UL trails by 38.1% with 1266 points against 2044. Every other test shows a larger gap. The 338H wins all three Cinebench generations in both single-core and multi-core, and it wins every Passmark workload: data compression, data encryption, extended instructions, prime number finding, floating-point math, integer math, multithread, physics, random string sorting, and single-thread. The use-case split is therefore not a split at all in performance terms. The 120UL's advantages are platform attributes rather than benchmark wins: it is a desktop Socket 1700 part, it supports DDR4 and DDR5, and it draws 15W, which is 10W lower than the 338H's 25W. Those attributes allow the 120UL to fit systems the 338H cannot, but they do not produce a single head-to-head victory.

Architecture Differences

The architecture gap is generational. The Core 5 120UL uses Raptor Lake, specifically the Raptor Lake-PS codename, on Intel's 10 nm process. The Core Ultra 5 338H uses Panther Lake, the Panther Lake codename, on Intel's 3 nm process. The 338H has 12 cores and 12 threads; the 120UL has 10 cores and 12 threads. Both expose 12 threads, so the 338H's additional two cores do not raise thread count. Base clocks are 1.90 GHz for the 338H and 1.30 GHz for the 120UL, while boost clocks are 4.70 GHz and 4.60 GHz. Cache sizes are substantially larger on the 338H: 192 KB of L1 per core versus 80 KB, 2.5 MB of L2 per core versus 1.25 MB, and 18 MB of shared L3 versus 12 MB. Memory support differs completely. The 120UL accepts DDR4 and DDR5 on a dual-channel bus. The 338H accepts LPDDR5X on a dual-channel bus and records 136.5 GB/s of memory bandwidth. PCIe connectivity also differs: the 120UL provides Gen 4 with 8 CPU lanes, while the 338H provides Gen 5 with 4 CPU lanes. Integrated graphics move from Iris Xe Graphics 80EU on the 120UL to Arc B370 on the 338H. The 120UL is a desktop part on Intel Socket 1700; the 338H is a mobile part on Intel BGA 2540. Release dates are 2024-04-07 for the 120UL and 2026-01-04 for the 338H. Neither processor has an unlocked multiplier, and neither supports ECC memory.

FAQ

Q: Which CPU wins in multi-core rendering?

A: The Core Ultra 5 338H wins every multi-core Cinebench test. In R23 multi-core it scores 16331 versus 8974, leaving the 120UL 45% behind. In R20 multi-core the gap is 63.1% (10213 versus 3769), and in R15 multi-core it is 63.9% (2504 versus 904).

Q: How large is the single-thread gap?

A: The Core Ultra 5 338H leads all single-thread tests. Cinebench R23 single-core shows 2044 versus 1266, a 38.1% deficit for the 120UL. Passmark single-thread shows 4180 versus 2080, a 50.2% deficit.

Q: Which processor supports DDR4?

A: Only the Core 5 120UL. Its memory support field lists DDR4 and DDR5, while the Core Ultra 5 338H lists LPDDR5X only.

Q: Which processor has more cache?

A: The Core Ultra 5 338H. It has 192 KB of L1 per core, 2.5 MB of L2 per core, and 18 MB of shared L3. The 120UL has 80 KB of L1 per core, 1.25 MB of L2 per core, and 12 MB of shared L3.

Q: Are the integrated graphics different?

A: Yes. The 120UL uses Iris Xe Graphics 80EU, while the 338H uses Arc B370.

Q: What are the market segments?

A: The 120UL is a Desktop part on Intel Socket 1700. The 338H is a Mobile part on Intel BGA 2540.

Specification Differences

| Field | Core 5 120UL | Core Ultra 5 338H |

| Series | null | Core Ultra Series 3 |

| Cores | 10 | 12 |

| Base clock | 1.30 GHz | 1.90 GHz |

| Boost clock | 4.60 GHz | 4.70 GHz |

| TDP | 15 W | 25 W |

| Socket | Intel Socket 1700 | Intel BGA 2540 |

| Architecture | Raptor Lake | Panther Lake |

| Codename | Raptor Lake-PS | Panther Lake |

| Generation | Core 5 (Raptor Lake-PS) | Ultra 5 (Panther Lake-H) |

| Process node | 10 nm | 3 nm |

| L1 cache | 80 KB (per core) | 192 KB (per core) |

| L2 cache | 1.25 MB (per core) | 2.5 MB (per core) |

| L3 cache | 12 MB (shared) | 18 MB (shared) |

| Memory support | DDR4, DDR5 | LPDDR5X |

| Memory bandwidth | null | 136.5 GB/s |

| PCIe | Gen 4, 8 Lanes (CPU only) | Gen 5, 4 Lanes (CPU only) |

| Integrated graphics | Iris Xe Graphics 80EU | Arc B370 |

| Market segment | Desktop | Mobile |

| Release date | 2024-04-07 | 2026-01-04 |

| Part number | unknown | SA4REQ9EW |

| Percentile vs all CPUs | 68 | 84 |

| Average benchmark score | 13594 | 33989 |

Head-to-Head Benchmarks

The head-to-head table contains 17 tests, and the Core Ultra 5 338H wins all 17. The largest margin is in Passmark find prime numbers: 304 for the 338H versus 47 for the 120UL, an 84.5% deficit. Passmark extended instructions shows a 78.2% deficit, with 23906 against 5203. Passmark physics follows at 70.1%, scoring 2697 versus 807. Floating-point math is 68.7% in favor of the 338H, 84067 to 26311. Data encryption lands at 64%, 21367 versus 7685, and Cinebench R15 multi-core is nearly identical at 63.9%, 2504 versus 904. Passmark multi-thread and Cinebench R20 multi-core sit at 63.2% and 63.1% respectively, with scores of 28717 versus 10558 and 10213 versus 3769. Cinebench R20 single-core also trails by 63.2%, 1441 versus 531. Cinebench R15 single-core trails by 58.4% (305 versus 127), data compression by 60.6% (276539 versus 109090), and random string sorting by 60.1% (34082 versus 13610). The smallest margins are still large: Cinebench R23 single-core has a 38.1% deficit (2044 versus 1266), and Passmark integer math has a 41.4% deficit (64934 versus 38060). Cinebench R23 multi-core falls in between at 45%, with 16331 against 8974. In every category, the 120UL trails by at least 38.1%, which indicates a consistent performance advantage rather than a workload-specific one.

The Verdict

The benchmark data points to one clear conclusion: the Core Ultra 5 338H is decisively faster than the Core 5 120UL. It wins all 17 head-to-head tests, posts an average benchmark score of 33989, and ranks in the 84th percentile of all CPUs. The 120UL averages 13594 and ranks in the 68th percentile. The 338H's nearest-rival group includes the Core Ultra 7 165H at a -0.3% delta and the Core i7-12800HX at 0.3%, which places it in a higher performance class than the 120UL, whose nearest rivals include the Core i3-12100F at 0.7% and the Core 3 N355 at 0.8%. The 120UL's only advantages are platform-level: it is a desktop Socket 1700 processor, it supports DDR4 and DDR5, and it has a 15W TDP versus 25W. On performance, the data indicates the Core Ultra 5 338H. On platform compatibility, the Core 5 120UL is the only option of the two for desktop Socket 1700, DDR4/DDR5 memory, and a 15W TDP envelope. The recorded data offers no performance scenario where the 120UL wins.